Johan Stokking
Johan Stokking
> Some thoughts: > > [1..5] > > I think (3) is more reasonable, the rest seem smelly / bad design. At least (3) is defensible in a way. I...
The last update of the repository Actions secret `SNAPCRAFT_LOGIN` is 4 months ago.
We can do it in three places; 1. Login with entering password. I think we should enforce it at least at this point, i.e. after the user typed their current...
> Also, since there is no `DeviceModeInd` in 1.0.x, how would a device signal to the NS the switch ? The application does that, so that would go via the...
If the culprit is indeed that GS's `ScheduleAnytime()` schedules downlink far in the future, then the cause could still be an already scheduled downlink that GS is scheduling after. Like...
> Edit: It is theoretically possible to use an old downlink path, via a token, to schedule a class C downlink via Packet Broker. It's a bit weird, but not...
> I start to wonder if allowing the delay to grow indefinitely is a good strategy. What happens now is that a downlink path may consume all of the downlinks...
Since we have users of this, let's do this as part of a major bump. It doesn't hurt too much keeping it in the codebase.
I think we can indeed validate (and remove) the LoRa RX metadata if the RSSI is outside of the pretty forgiving range [-142, -15] and SNR [-20, 15]. We can...